PROSURProgram for the study of regional climate variability,their prediction and impacts, in the Mercosur area

Report on

CLIMATE VARIABILITY IN INTERDECADAL, INTERANUAL AND INTRASEASONAL TIME SCALES: IMPACTS AND MECHANISMS

Rapporteurs: Alice GrimmTercio Ambrizzi

3th Co-PI's Meeting Mar del Plata October 14-17, 2002

Berri and Bertossa: The influence of tropical and subtropical Atlantic and Pacific Oceans on precipitation over southern-central South America (SCSA) (submitted to J. Climate)

CCA of bi-monthly precipitation and SST, using 10 EOFs of SST and 9-11 EOFs of precipitation, that explain at least 80% of the variance.

The strongest association between SST and precipitation: November-December, followed by Mach-April and May-June.

The weakest association: January-February.

 Identification of oceanic regions and SCSA whose SST and precipitation anomalies are correlated. Example of November-December: the Pacific Ocean has stronger and more spread out influence over SCSA than the Atlantic Ocean.

 Relationship with - tropical oceans: warm-wet, cold-dry- subtropical oceans: warm-dry, cold-wet.

SWPSP

N 3N 3.4

TP

NWP

SEP

SA

TSA

EA

NA

SELECTED OCEANIC REGIONS

SWPSP

N 3N 3.4

TP

NWP

SEP

SA

TSA

EA

NA

SWPSP

N 3N 3.4

TP

NWP

SEP

SA

TSA

EA

NA

SELECTED OCEANIC REGIONS

SWP

TP, NA,

N 3, N 3.4

SA TSA,

EA SWP

SP,SWPN 3.4

TP, SEP, NA

N 3, N 3.4

TP, SEP,

SWP, SP

N 3, N 3.4

TP SP

N 3.4

SWP

N 3, N 3.4

TP, SEP,

SP, SWPTP, SEP

SWP

N 3, N 3.4

TP, SEP SP, SWP

N 3, N 3.4 TP, SEP, NA

SWP

REGIONAL SST INFLUENCE

NOVEMBER – DECEMBER (1961-1993)

SWP

TP, NA,

N 3, N 3.4

SA TSA,

EA SWP

SP,SWPN 3.4

TP, SEP, NA

N 3, N 3.4

TP, SEP,

SWP, SP

N 3, N 3.4

TP SP

N 3.4

SWP

N 3, N 3.4

TP, SEP,

SP, SWPTP, SEP

SWP

N 3, N 3.4

TP, SEP SP, SWP

N 3, N 3.4 TP, SEP, NA

SWP

REGIONAL SST INFLUENCE

NOVEMBER – DECEMBER (1961-1993)

Barros and Silvestri: The relation between SST at subtropical south-central Pacific and precipitation in southeastern South America (SSA). (J. Climate, 2002)

The two first modes of CCA (austral spring) show warm-wet and cold-dry relationship between tropical SST and precipitation.

SST in the equatorial regions does not modulate rainfall variability among EN and LN cases, but between EN and LN cases in the whole period and among neutral cases.

The SSA rainfall response to equatorial Pacific SST is not linear.

SST in the SSCP modulates the seasonal rainfall over most of SSA.

SST in Niño 3 and Niño 3.4 and SST in SSCP are negatively correlated over the whole period, but are uncorrelated for only EN or neutral cases.

Stratification of cases: difference between cases with constant equatorial SST and different SST in SSCP shows a ENSO-like wave-train. Most of the ENSO signal at mid and high latitudes is associated with SST variability in SSCP.

(a) Precipitation stations used in this study and the regions considered. (b) Ocean regions with the highest SST correlation with precipitation over SSA in OND. The shaded area in Equatorial Pacific is the Niño 3.4 region.

(a)

(b)

TABLE 1: Correlation between precipitation over SSA regions and SST at the six ocean regions depicted in Fig. 1 for different periods. Significant values at 95 % level in bold.

FIGURE 2: Scatter diagram between OND precipitation in region NE and SST in ocean regions as indicated in each panel. El Niño (dots) and La Niña (empty squares) events, and neutral cases (crosses). Correlations (r) for all cases, only EN cases (dots), LN cases (empty squares) and neutral cases (crosses) are indicated below each panel.

Vera, Silvestri, Barros, and Carril: Inter El Niño and La Niña variability of the Southern Hemisphere circulation during austral spring. (submitted to J. Climate)

Stratification of the springs associated with ENSO events according to SST conditions over El Niño 3.4 sector and over the SSCP region: WC and WW cases. 

WC cases: enhanced convection not only in the equatorial central Pacific but also in subtropical regions of the southeastern Pacific Ocean. Intensificaton of a localized Hadley produces a well-defined Rossby wave pattern in the central south Pacific extratropical region.

WW cases: anomalous circulation pattern more concentrated in the tropics.

The heating forcing present in the WC cases in the south central Pacific subtropical region is essential to maintain the wavelike circulation over that area.

Composite of SST and streamfunction (at =0.2) anomalies for WC and WW

El Niño events.

WC

WW

Grimm:The El Niño impact on summer monsoon in Brazil: regional processes versus remote influences. (J. Climate, in press)

Grimm: How do La Niña events disturb the summer monsoon system in Brazil? (submitted to Climate Dynamics)

Data: precipitation from more than 1800 selected stations, observed surface temperature, and other atmospheric data from reanalysis.

The El Niño (La Niña) impact on Brazil summer monsoon is not well assessed through seasonal analysis because of subseasonal variations. Precipitation and circulation anomalies that are consistent during part of the season, are smoothed out in a seasonal analysis.

There are abrupt changes of anomalies within the summer monsoon season, suggesting the prevalence of regional processes over remote influences during part of the season.

In early season remotely produced atmospheric perturbations prevail over Brazil. Perturbations in the Walker and Hadley circulations over East Pacific and South America, and a Rossby wave-train over southern SA favor negative (positive) precipitation anomalies in north and central-east Brazil and positive (negative) ones in south Brazil.

In January, with the enhancement (weakening) of the continental subtropical heat low by anomalous surface heating during the spring, there is anomalous low-level convergence (divergence) and cyclonic (anticyclonic) circulation over southeast Brazil, while at upper-levels anomalies of divergence (convergence) and anticyclonic (cyclonic) circulation prevail.

This anomalous circulation directs moisture flux towards central-east Brazil (south Brazil), causing moisture convergence in this region. A favorable thermodynamic structure enhances precipitation over central-east Brazil (south Brazil), the dry (wet) anomalies in north Brazil are displaced northward, the wet (dry) anomalies in south Brazil disappear.

In February there is a return to previous conditions.

Influence functions indicate that while the anomalies of circulation over southeast Brazil in spring of El Niño years can be attributed to remote influences from the tropical East Pacific, those in January are probably due to local influence. In this month the monsoon-like circulation is enhanced.

Simultaneous and lagged correlation analysis of SST and rainfall in central-east Brazil shows that SST anomalies in the Atlantic Ocean off the southeastern coast of Brazil fluctuate on the same timescale as the circulation and precipitation anomalies.

EL NIÑO LA NIÑA

Precipitation percentiles expected for El Niño/La Niña events. Shadowed areas have precipitation anomalies consistent over

90% (85%) confidence level.

Moisture fluxEL NIÑO LA NIÑA

Composite of vertically integrated moisture flux anomalies (left) for the indicated month of El Niño events. Units are 10-2mgs-1kg-1

and 10-2gs-1kg-1, respectively. The shaded areas indicate anomalies consistent over 90% confidence level.

Thermodynamic Structure

EL NIÑO LA NIÑA

Composite profiles of the anomalies ’, e’ and es’ for the indicated month of El Niño events, calculated from the reanalysis data averaged over the region 15S – 22.5S, 40W – 47.5W, in Southeast Brazil. Units are Kelvin.

Influence Functions Analysis

Influence functions for a target point in Southeast Brazil indicated by a black circle, calculated for November, December and January basic states.

Souza e Ambrizzi: Further analysis of ENSO impacts on the subtropical South America (SSA): evolution of the precipitation anomalies on pentad timescale. (To be submitted)

Composites of pentad GPCP precipitation and 200 hPa wind anomalies for El Niño and La Niña events during the period 1979-2000.

Positive (negative) precipitation anomalies are stronger in South Brazil.

The precipitation anomalies are rather uniform during El Niño (La Niña) episodes, but there is a reversal of sign over part of the region in periods of 1 to 3 pentads.

The strengthening (weakening) of the subtropical jet over SSA is associated with positive (negative) precipitation anomalies in the region.

Composite pentad precipitation anomalies (shaded) and magnitude of the 200 hPa wind (green contours with m/s units) associated with the El Niño episodes. Precipitation anomalies (mm/day) are departures from the 1979-1999 long-term-mean. The shading precipitation interval are shown in the colorbar. The magnitude of the wind is plotted above 20 m/s with interval of 5 units.

Antico e Berri: Possible mechanisms responsible for strengthening and weakening the monthly mean subtropical jet over South America during the austral winter. (To be submitted)

Composites for 4 cases of strong STJ and 5 cases of weak STJ in July-September of the period 1978-1999.

Strong (weak) STJ is due to strong (weak) Hadley cell associated with enhanced (suppressed) convection in the eastern Pacific, off the Central America coast.

There are 2 cases different from the other ones: a strong STJ in 1995 and a weak one in 1992, due to Rossby wave propagation from the western Pacific (1995) and from tropical South America (1992).

Anomalous STJ produces anomalous meridional circulation downstream (indirect meridional cell) and consequently upward motion and cloudiness.

STJ

Difference between composites of OLR anomalies of intense and weak STJ cases. Solid (dashed) lines represent positive (negative) values in W m – 2. Shaded areas indicate regions with statistically significance equal or greater than 5%. The period is July-September 1978-1999.

Pastalanga, Vera, and Piola: On the leading modes of the sea surface temperature variability in the South Atlantic.(Clivar Exchanges, 2002)

EOF analysis of the monthly mean SST fields from the NCEP-NCAR reanalysis, for 1972-2000.

S-EOFs are obtained from analysis in the spatial domain and T-EOFs are obtained from analysis in the time domain.

T-EOF1: north-south dipole with dominant interdecadal variability

T-EOF2: strong center in subtropical central Atlantic with centers of opposite sign south of Africa and off the Argentinean coast. Interannual and sub-annual variability (located over the Brazil-Malvinas confluence).

T-EOF3: east-west dipole at mid-latitudes with significant interannual variability associated with ENSO.

Only the first S-EOF mode agrees with a corresponding T-EOF mode.

The EOF analysis in the time domain is an efficient tool to isolate the SST leading patterns without requiring additional filtering.

(Left panels) principal components of the first three modes of the T-EOF analysis. Contour interval: 0.3°C. (Right panels) spectra of the first three T-EOFs including red noise spectra and 95% confidence bands.

Grimm and Natori: Relationship between the interannual and interdecadal variability of precipitation in southeastern South America and SST in the Pacific and Atlantic oceans. (presented at the XII Brazilian Meteorological Congress; to be submitted).

CCA of rainfall in 515 stations from northeast Brazil to Patagonia and SST (HADISST) from Pacific and Atlantic oceans in austral spring.

The analysis for the Pacific shows 3 ENSO-like modes (1, 2 and 5) with differences in the location of the largest SST anomalies, tropics-extratropics SST gradients, and precipitation anomalies.

The analysis for the Atlantic shows 2 modes (1 and 3) correlated with Niño 3 SST, and correlated with the three ENSO-like modes of the Pacific. Mode 4 is not correlated with Niño 3, but is correlated with mode 1-Pacific.

Those different modes have different interdecadal oscillations, which explain the interdecadal modulation of El Niño events.

In the analysis for both oceans modes for each one of them are combined in some cases, indicating joint oscillations. Sometimes their effect on precipitation is the same, but sometimes it is opposite. Sometimes the isolated influence of one ocean is more important.

Mode 1 – Pacific

Mode 2 – Pacific

Mode 1 - Atlantic Mode 3 - Atlantic Mode 4 - Atlantic

Mode 2 – Pacific+Atlantic

This mode is a mixture of the modes 1 and 2 for Pacific, and modes 3 and 4 for the Atlantic (which shows up in the canonical map for precipitation).

Mode 4 - Atlantic Mode 3 - Atlantic Mode 1 - Atlantic

Correlation between the SST canonical vector and SLP

Nobre and Obregón: Linear tendency and interdecadal variability of the precipitation over South America (based on the thesis of Obregón, to be submitted)

The linear tendency and interdecadal variability of annual precipitation are determined with non-parametric statistical methods for 1951-1990.

There are different regional tendencies: negative over northwestern Amazonia and southeast Brazil, and positive ones in south Brazil and the subtropical region east of the Andes.

There are abrupt changes in the 70´s, apparently associated with ENSO-like interdecadal variability.

Linear tendency of the annual precipitation with respect to the 1951-1990 mean. Tendencies significant at 95% level in the Mann-Kendall test are indicated by .

Annual precipitation (blue line), series smoothed by the method of local regression (red line), and linear tendency by the method of Sen (black line).

NW Amazonia

NE Brazil

SE Brazil

South Brazil

East of Andes